ASML is investing €1.3 billion ($1.5 billion) in Mistral’s €1.7 billion ($2 billion) Series C raise and is expected to secure a seat on the company’s board, people familiar with the matter told Reuters.
The deal values Mistral at €10 billion ($11.7 billion), making it Europe’s most valuable AI startup.
ASML declined to comment, while Mistral did not respond to requests for comment.
Mistral, often described as Europe’s AI champion, competes with US players such as OpenAI and Google. Founded in 2023 by Arthur Mensch, a former DeepMind researcher, and ex-Meta researchers Timothée Lacroix and Guillaume Lample, it has been backed by Nvidia and other investors.
Sources said ASML’s investment ties together two leading European tech firms and could reduce the region’s reliance on US and Chinese AI models.
ASML supplies extreme ultraviolet lithography equipment, which is critical for producing the most advanced semiconductors. The systems cost about $180 million each and are used by chipmakers such as Taiwan Semiconductor Manufacturing Co and Intel.
The company also applies AI in its own tools to improve performance. Access to Mistral’s data analytics and machine learning models could enhance ASML’s efficiency and help develop new products.
Mistral was valued above $6 billion after its Series B round last year. Earlier this month, the Financial Times reported the company was in talks to raise $1 billion at a $10 billion valuation. Bloomberg separately reported the new funding round could value the startup as high as $14 billion.
Bank of America is advising ASML on the investment, the people said. The bank declined to comment.
In recent years, European leaders have expressed concern over US dominance in artificial intelligence. By supporting startups such as Mistral, European governments aim to prevent the United States from becoming the sole power shaping the future of the field. They argue that the concentration of technology in US hands threatens Europe’s scientific and economic independence, while full reliance on American models risks what they describe as a new form of digital colonialism.